Where should this person be, so the machine cannot do what they do?
Deployment as the first talent decision. Placing a person where their judgment is the throughput, not their tasks.
The talent choices for the AI transition.
If Judgment Call is about the manager's remaining edge, and Lead What Lasts is about the organizations that make good judgment possible, then Unreplaceable is about the people. Who to hire, who to keep, and how to make sure the person the machine cannot substitute for is the one on your team.
Book three of the More Capable series. The talent choices that make a person hard to substitute for, and how a manager designs a team of them.
Every wave of automation has ended the same way. The tasks the machine can do become the tasks the machine does. The people who kept their jobs, or grew into new ones, were the ones whose real contribution was never the task at all. It was the judgment, the relationships, the initiative, the care, or the pattern-sensing the machine could not substitute for. That is the pattern this book takes seriously.
The mistake most managers are making in 2026 is treating "AI-proof" as a role description. It is not. It is a talent-development choice, made across the four paired activities of the DART model: how you Direct and Deploy your people, Attract and Assess them, Reward and Retain them, Teach and Transform them over time. Get the four right, and you compound capability. Get them wrong, and you build a team that is technically fast and structurally fragile.
Unreplaceable is not a personality type. It is a talent design.
Unreplaceable is in early development. The framework is in place; the drafting is underway. Sign up below and I will let you know when the pre-order goes live.
Eight questions the book will answer, one for each paired activity in the DART model plus the two that hold the framework together.
Deployment as the first talent decision. Placing a person where their judgment is the throughput, not their tasks.
The interview questions, portfolio reviews, and reference checks that surface the qualities the AI cannot replicate.
Judgment, dignity, initiative, and the qualities that show up in how a person handles a hard case, not a resume line.
Compensation as a design language. Rewarding the behaviors that make someone unreplaceable rather than the tasks they complete.
The unwritten contract with the people whose judgment holds the org together, and how to make it more explicit.
Apprenticeship, feedback, and the reps that build the capability the AI cannot supply.
Transformation as a management responsibility. Helping your people grow through the change rather than around it.
The closing chapter. The five moves that keep an unreplaceable person from becoming someone else's unreplaceable person.
One email when Unreplaceable is available for pre-order. No newsletter, no drip, no roundup. Just the message.
Or subscribe to the More Capable newsletter for one idea a week between now and launch.Thomas J. Norman, Ph.D. is a Professor of Management at California State University, Dominguez Hills, and a Gallup-certified executive coach. He teaches, researches, and consults on how organizations turn human dignity into measurable results.
His executive coaching practice serves CEO, COO, CFO, and CHRO leaders integrating AI into growing companies. The talent decisions he coaches them through every week are the material this book takes up in full.
He writes the weekly newsletter More Capable and hosts the interview series Off the Talking Points from the Organizational Leadership Effectiveness Lab in San Pedro.
Three books, one argument. As AI absorbs routine cognitive work, the manager's real job becomes growing talent that gets more capable, not more dependent.
How great managers decide what AI cannot.
The organizations that make good judgment possible.
The talent choices for the AI transition.